Welcome to Founder Frames.
Where we decode the systems, strategies, and business mechanics of underrepresented founders.
The Frame: The Inverse Scaling Method
How Elias Torres is making the boldest bet in enterprise software by building a billion-dollar company with fewer than 100 employees.
This is a masterclass in operational leverage and a blueprint for founders who refuse to play by the old rules.
The Customer Success Ceiling
Here's a math problem killing SaaS companies:
The average Customer Success Manager (CSM) can handle 50-100 accounts.
To manage 10,000 customers, you need 100-200 CSMs.
At $80K average salary plus overhead, that's $12-24 million in annual costs before you've delivered a single outcome.
But here's the trap nobody talks about: even with those costs, most CSMs spend 70% of their time on repetitive tasks.
Sending follow-up emails
Scheduling check-ins
Preparing for calls
Logging notes
The industry's solution? Hire more people. Burn more cash. Pray for efficiency gains.
Torres had a different insight: The problem isn't that humans can't scale. It's that we're scaling the wrong work.
THE INVERSE SCALING METHOD
Most founders think scale means: more customers → more employees → more complexity.
Torres inverted this entirely.
His new company, Agency, is built on three radical principles:
Automate the 70%, Elevate the 30%
Instead of hiring CSMs to "do more with less," Torres built an AI agent called Kai.
Kai handles the entire 70% of administrative work—scheduling, prep, follow-ups, onboarding sequences, and data entry.
The key insight? Kai doesn't replace humans. It frees them to do what only humans can: build deep relationships, solve complex problems, and drive strategic value.
The 100-Employee Ceiling
Torres has publicly committed to reaching $1 billion in revenue with no more than 100 employees.
Today, Agency has - 15 people, mostly engineers.
Why the constraint?
Because constraints drive innovation.
Every hire forces you to ask: Can AI do this instead?
Every process gets stress-tested for automation potential before adding headcount.
This is architecture by limitation - building systems that scale without human dependency.
Bet on the Paradigm Shift
Torres consulted with OpenAI in 2022 and saw what was coming.
His thesis: the companies that win the next decade aren't those who "add AI" to their existing models.
They're the ones who rebuild their entire operating system around what AI makes possible.
Agency raised $12M from Sequoia Capital and HubSpot Ventures, then $20M Series A from Menlo Ventures.
Heavy hitters betting that Torres is right about the future of work.
THE 15-PERSON OPERATING MODEL
How does a company with 15 people deliver enterprise-grade customer success?
There's no separate "Customer Success team," "Sales team," or "Support team."
There are engineers and AI agents that handle everything else.
When a customer signs up, Kai handles onboarding automatically.
It ingests the customer's CRM data and historical support tickets.
It builds a personalized success plan within 24 hours.
It schedules the first human touchpoint only when strategic value requires it.
The metric Torres tracks: Time-to-first-value (TTFV).
Traditional onboarding takes 14+ days. Kai brings it under 48 hours.
Agency's 15-person team doesn't get every customer interaction.
Kai handles 80%+ of inbound requests autonomously.
Humans are triggered only when:
The dollar value at risk exceeds $50K
The customer explicitly requests human intervention
Kai's confidence score drops below 70%
This isn't chatbot triage.
Kai actually executes tasks, rescheduling calls, generating QBR decks, even drafting contract amendments for legal review.
Torres's north star: 10x the output of a traditional SaaS team.
If any function starts requiring headcount, the first question is always: Why can't Kai do this?
THE FRAME
Torres's playbook reveals three principles any founder can apply:
1. The Constraint Architecture
Set a hard limit before you need it.
Torres's 100-employee ceiling forces every decision through a filter: Does this scale without human dependency?
Torres chose 100 employees. You might choose $100K revenue per head.
The number matters less than the forcing function.
2. The 70/30 Audit
Map where your team spends their time.
What's the 70% of repetitive work that can be systematized or automated?
What's the 30% that's genuinely high-value?
Tomorrow's action: For your highest-cost team, have each member log their tasks for one week. Categorize: "Repetitive" vs. "Strategic."
Anything in the repetitive bucket with >4 hours/week is your automation target list.
3. The Paradigm Bet
The Drift playbook was a paradigm shift in 2015.
Torres isn't rerunning that playbook.
He's betting on the next shift: agentic AI.
Tomorrow's action: Finish this sentence and put it in your notes:
"In 3 years, [MY INDUSTRY] will be transformed by [PARADIGM SHIFT]. My company will be built for that world because we're [SPECIFIC DECISION] today."
THE BOTTOM LINE
Elias Torres isn't just building an AI company.
He's running an experiment: can you 10x a business while 10x-shrinking the team required to run it?
He's choosing the hardest possible path because he believes it's the future.
You don't play for safety.
You play to change the rules.
Until next Thursday,
AP